On August 16, 2025, the Nepal government rolled out its first National Artificial Wisdom (AI) policy, which was taken as "ambitious" and "visionary." But as a policy practitioner, I see this policy becoming a decorative strategy—light on promise and light on applied action. Unless it is seriously followed through, it won’t fix the inequalities it talks about; instead, it could make them worse.
Globally, AI is projected to add $15.7 trillion to the world economy by 2030. Nepal, with a GDP of just $44 billion, cannot afford to sit idle. Yet we must ask: does our current institutional capacity allow us to meaningfully participate in this technological shift, or are we simply drafting policy to signal modernity without the muscle to implement it?
The policy speaks at length about human capital development, but it separates from reality. Every year, Nepal produces around 50,000 ICT graduates, but less than 20 percent are considered employable in advanced technologies such as AI. Where is the roadmap for the course of Olhal University, encouraging AI-centered research, and retaining talent that is fleeing abroad? Without structural reforms in higher education, this policy will produce eloquent documents, not competent engineers.
The government also proposes a National AI Centre and an AI Regulation Council. On paper, this sounds impressive. But let us not forget: Nepal’s regulatory bodies, from the Telecommunications Authority to the Cyber Bureau, already struggle with enforcement. Adding new councils without clarity on authority and accountability only bloats bureaucracy. Who will actually protect citizens when deepfakes undermine elections, or when AI-driven loan algorithms discriminate against women and Dalits? The policy provides no answers.
Building data centres in the Himalayas sounds like a smart idea. Cold weather means lower cooling costs. But let’s be honest: who will actually invest when we still face daily power cuts, shaky internet connections, and unstable politics? Dreaming of cloud infrastructure is easy. Paying the electricity bill on time is the real challenge.
Will Covid-19 make us humane?
Startups are another buzzword in the document. According to FNCCI, Nepal has around 300 tech startups, but fewer than 10 percent receive venture funding. Compare this with Vietnam, which secured over $1 billion in startup investment in 2022. Without serious reforms to taxation, foreign investment rules, and bureaucratic red tape, what investor will take Nepal’s AI rhetoric seriously? Our brightest young entrepreneurs will continue to migrate to Bangalore, Singapore, or California, while “support for startups” remains a bullet point in government memos.
However, the most spectacular weakness is equity and inclusion. Women make only 24 percent of Nepal's ICT workforce. If this policy does not actively embed inclusion targets—scholarships, gender audits of AI systems, and encouragement for women-led startups—it will repeat the bias of the offline world in digital. AI does not erase inequality; it amplifies it. Through the Project Executive, I have noticed how quickly women and marginalized youth can be excluded from new systems if digital literacy is not a priority. We have worked with 4,500 young women in 47 districts, giving them equipment to engage politically and digitally. Imagine what happens when AI starts shaping regime, health, or finance—if these similar communities are left behind.
The government promises a two-year review of the policy. Reviews without enforcement are meaningless. Nepal’s history with ICT and e-governance policies is littered with half-finished initiatives and stalled implementation. Unless the state allows genuine public-private partnerships and independent oversight, this AI policy risks becoming just another addition to the archives.
Nepal does not need a symbolic AI policy—it needs a grounded one. That means:
* Restructuring higher education to build real AI capacity.
* Fixing electricity and internet infrastructure before chasing cloud dreams.
* Reforming tax and investment rules to retain startups.
* Setting measurable gender and inclusion benchmarks, not vague aspirations.
* Establishing grievance redressal mechanisms for citizens affected by AI harms.
AI is not neutral. It is shaped by who designs it, who funds it, and who regulates it. If Nepal’s AI policy sidelines its people—especially women, rural youth, and marginalized groups—it will deepen inequality under the guise of innovation. If, however, we confront these gaps head-on, Nepal can carve out a unique space as a regional leader in ethical and inclusive AI.
The question is blunt: will this policy empower citizens, or simply flatter elites? Nepal cannot afford another “ambitious” strategy that collapses at the first test of implementation.
The author is the Founder and Chairperson of Project Abhaya, a youth-led initiative advancing gender-inclusive political and digital literacy in Nepal. She leverages AI to improve governance, streamline policy-making, and maintain data on Members of Parliament, while developing ethical and inclusive frameworks for digital policymaking.
